Colorectal Cancer Podcast

When Should You Consider Immunotherapy in Colorectal Cancer?

Benjamin L. Schlechter, MD; Marwan Fakih, MD


June 01, 2023

This transcript has been edited for clarity. For more episodes, download the Medscape app or subscribe to the podcast on Apple Podcasts, Spotify, or your preferred podcast provider.

Benjamin L. Schlechter, MD: Hello. I'm Dr Benjamin Schlechter and welcome to Medscape's InDiscussion series on colorectal cancer. Today we're discussing immunotherapy for colorectal cancer, and Dr Marwan Fakih is joining us. Dr Fakih is a professor in the Department of Medical Oncology and Therapeutics Research and co-directs the Gastrointestinal Cancer Program as the Judy and Bernard Briskin Distinguished Director of Clinical Research at City of Hope. Dr Fakih, thank you for joining us.

Marwan Fakih, MD: Thank you for having me.

Schlechter: It's my pleasure. You've had an impressive and distinguished research career in gastrointestinal (GI) oncology, and more recently, you've led several important innovations in immunotherapy in colorectal cancer, which is really exciting because this has been a cold tumor for a long time. Before we move to that topic, I'd love to hear a bit more about your path to becoming a GI oncologist and a successful investigator in this field.

Fakih: This is one area where I can say I did not really plan the path, to be honest with you. I've always, even as a medical student, had a passion for oncology. Initially, frankly, I was more interested in hematology than oncology. As I went through fellowship, I went toward the solid tumor route, and I was always interested in the diversity, complexity, integration of different specialties within oncology. I would have to say my career moved toward GI oncology because I remained at my mother institute — University of Pittsburgh — as an assistant professor of oncology. The opportunity came up to work in phase 1 trials in GI cancers at that point, and I fell in love with colorectal cancer. From there, I kept steady at it and enjoyed every minute of it. Sometimes chance moves you in one direction, and for me that was a good direction.

Schlechter: Great, and it's been a successful one. Let's start with microsatellite-unstable colorectal cancer. This has been, I think, the real success story of checkpoint blockade in GI cancers. Let's start at the bottom, and we'll move our way up as a way of organizing things.

About a year ago, we all heard the impressive data published in The New England Journal of Medicine, by Dr Andrea Cercek from Sloan Kettering and colleagues, using dostarlimab for localized mismatch repair–deficient rectal cancer. This was a small phase 2 study, but many call this a practice-changing result. I'm curious to hear your thoughts on these data and how you now treat mismatch repair–deficient localized rectal cancer based on this information.

Fakih: What a success story, right? Between that trial and the NICHE trial looking at that microsatellite instability (MSI)–high colorectal cancer with localized disease, we know that you don't need many doses of immunotherapy to get the tumor to shrink away. I think although all of us love to see confirmation with larger trials, the reality is that patients with rectal cancer, with MSI-high disease, do need better treatments. When we think about immunotherapy and sphincter saving — things like avoiding surgery, avoiding radiation therapy — many of us ask, do we want to wait another five years to offer those strategies to our patients? The answer is, in my opinion, no. I think we have to have that discussion. A 20-patient trial, more or less, with 100% response rate, coupled with supportive data from the NICHE trial showing a very high complete pathologic response with only 4 weeks of therapy prior to surgery, can't be really ignored.

I think our practice has changed. We don't start chemoradiation or chemotherapy without knowing that MSI status for localized rectal cancer, because we know that the discussion related to immunotherapy should be on the table with those patients. Whether we decide at that point to use this off-label or not is a different question.

Schlechter: Rectal cancer does seem like a place where it seems clear we should move forward. What are your thoughts on the NICHE study? Are you incorporating that into regular practice or for a laparoscopic resection, where maybe the morbidity of surgery and the outcomes are better? Are you routinely offering early neoadjuvant checkpoint inhibitor therapy for those patients, or are you still thinking in a more conventional resection-adjuvant chemotherapy pattern there?

Fakih: The reality is that patients with rectal cancer with MSI-high status are very rare and far in between. I think it's no more than probably 2% of our general population because they tend to be more the Lynch syndrome patients rather than those with MSI-high sporadic hypermethylation. When I see that rare instance with rectal cancer now, I do discuss upfront immunotherapy in those patients, and it would be an off-label discussion, obviously.

I don't think that I am at a point where I would suggest that for patients with colon cancer, I think even though the NICHE trial is extremely exciting, but the reality is the data is with a doublet. There is the issue of a lower-morbidity surgery in right-sided colon cancer, where you recover quickly from a laparoscopic right hemicolectomy and you avoid, even though rare, the potential significant immuno-oncology (IO) toxicities in some of those patients, and don't forget the need for more frequent endoscopic evaluations if you decide to proceed with immunotherapy only for colon cancer. I don't think we're there yet to offer patients with colon cancer immunotherapy as the only treatment modality for localized colon cancer. Rectal cancer may be different for me — and in that setting, it's a life-changing experience when you go with chemoradiation, and surgery, but unfortunately not for the good, right? We think about low anterior resection syndrome, lifelong bowel movement changes, urgency, frequency, etc.

Schlechter: It's an important distinction, right? Rectal cancer with the symptomology of treatment is very different than colon cancer. And the surveillance is different and harder.

You touched on single-agent checkpoint inhibitors vs doublets and the differences between them. That makes me think about advanced disease. When we think about mismatch repair–deficient advanced colorectal cancer, we're really first at checkpoint inhibitors. How do you balance the risks and benefits of, say, pembrolizumab vs ipilimumab-nivolumab (ipi-nivo), and when do you reach for one vs the other?

Fakih: Like many of us, I struggle with that decision because I don't think there is a right answer. We don't have today a definitive biomarker that tells us who is a patient who's less likely to respond to immunotherapy in MSI-high metastatic colorectal cancer, although there are some leads. For example, there are some data suggesting that diffuse carcinomatosis with ascites may be associated with a lower response to immunotherapy. We have previously shown that tumor mutational burden (TMB) within MSI-high colorectal cancer may be indicative of a response. When we see patients with a TMB of 25 or 28 or 19 with MSI-high, those are less likely to achieve a complete response, even an objective response. If I see a TMB of 30 or 25, I personally would rather go with a doublet, based on some of the data we published previously. Obviously if I have somebody with multiple organ involvement, there's some suggestion that liver metastases may have a little bit lower response, not just peritoneal metastases. In that setting, I may still favor a doublet in a younger individual with bulky disease.

I have to say this is not scientifically supported today by a wealth of data. That's just a pattern of practice that I have undertaken. I certainly have had many, many patients who've had a complete response and came off pembrolizumab after completing sometimes 8 months, 10 months, or even 1 year of treatment with pembrolizumab and not necessarily 2 years. They are years out now without a relapse, so overtreatment is still on my mind. I still find myself using more single agent pembrolizumab based on the phase 3 KEYNOTE-177 rather than proceeding with nivo-ipi in the majority of my patients.

Schlechter: And how do you surveil your patients when they're on treatment or off treatment, if they have no measurable disease? I always really, really struggle with that process.

Fakih: It's even more complicated because you may have measurable disease that is completely fibrotic and dead. We may be overtreating some patients if we rely completely on CT scans because we have many patients whom we start on pembrolizumab, and they'll have a liver metastases. We take them to surgery, and it's a complete pathologic response even though they do have visible, measurable disease on CT scans. So number one, you have to be aware as a provider that a CT scan underestimates the complete response to therapy. In my opinion and based on what we see in our practice, the reported complete responses in the literature of 10%-20% with pembrolizumab monotherapy are probably an underestimation of the complete pathologic response rate in these patients.

We rely somewhat on circulating tumor DNA (ctDNA) in our patients, and that's a very good biomarker of response to immunotherapy to make sure that you're not missing pseudo-progression early on. So we tend to get ctDNA at baseline and monitor, and if somebody has a complete ctDNA response and has an excellent response to imaging. If there's still a little residual disease on imaging, sometimes we rely on a PET scan to make sure whether this is metabolically active or nonmetabolically active. Even in someone with a little residual measurable disease with a PET-negative status or ctDNA-negative status, I may consider those more as a clinical complete response, in my opinion.

I would say as far as monitoring a complete response, I do scans every 3 months — CT typically if the patient doesn't have any measurable disease. I do that for a couple of years, and then after that I do it every 6 months and I monitor for years to come. I don't know that the 5-year mark that we're used to with stage III disease is the right thing here. After 5 years, I tend to do at least a yearly scan on those patients. I want to monitor them a little bit longer.

Schlechter: It's funny, without data we all seem to end in a similar place. I have this handful of patients with the same deal, where we're beyond 5 years. I remain nervous. One of them has what looks like measurable disease, but it's maybe a fibrotic node. And I still feel obligated to scan her every year. But we're only 7 years in, so I don't know. We'll see what I think at 10 years, but it's a real challenge. It's an exciting problem to have.

Checkpoint inhibitor therapy and immune therapy for mismatch repair–deficient colorectal cancer is critically important. As you pointed out, rectal cancer is quite rare and advanced disease is also quite rare. The elephant in the room in immune therapy for colorectal cancer has been microsatellite-stable disease, especially microsatellite-stable advanced disease. I thought we could move to that. I think that's a very exciting topic and a topic where you've really had a leadership role throughout the world on this particular approach.

The road traveled by checkpoint inhibitors has been bumpy and largely ineffective. We've had failure of ipi-nivo and durvalumab-tremelimumab. The FOLFOX and bevacizumab-nivolumab first-line study was a negative study. Even the GI002 neoadjuvant rectal cancer study, I think by and large we could call negative. The list is long, but now we're seeing some successes.

I want to start with the ipilimumab-nivolumab-regorafenib combo that you recently published. How did you come to this? What would you like us to know about this combination? How do you think about this? And maybe can you talk a little about the synergy of tyrosine kinase inhibitors (TKIs) and immune checkpoint inhibitors? Is it real? Do we believe it? I'd love to hear your thoughts.

Fakih: My interest was sparked initially, like all of us, by the Japanese data on regorafenib-nivolumab (rego-nivo) that were published in the Journal of Clinical Oncology, which at that time was like a 35% response rate. I have to say, like many other investigators across the country, we've used it off-label. One thing that I've noticed as I've treated my patients is that I've never seen a liver met respond. But I was seeing some patients with lung metastases develop responses to rego-nivo.

We conducted the phase 2 trial. Bayer supported that large phase 2 trial, and I served as a principal investigator to study rego-nivo in the US. The response rates were less than 10% in the overall population. It confirmed what we had previously reported in this series from our institute: that liver mets respond. There were more than 40 patients who did not respond, who had zero response rate with liver mets, and the response rates and patients without liver metastases was 22% with rego-nivo.

Now, I do think there is synergy there because as you said, the durvalumab-tremelimumab trial was completely negative. There was one responder on that study, and it was a lung met patient. But still, no more than 70% of patients in that large study had liver metastases and yet the responses were not 20% in those patients. Previously, with all the studies, with pembro, with nivo, we don't see that big signal with programmed cell death protein 1 (PD-1). Maybe you'll see some stabilizations, but not responses.

So I am convinced that the TKIs do add something, and we do see a little bit more flare, with skin toxicity that is not explained with rego alone, as if there is some kind of immune activation early on. Interestingly, when we looked at rego and nivo, there was some correlation between that particular toxicity and a little bit higher likelihood of response. On the basis of the rego-nivo data, we said, there was a signal, but how can we accentuate that signal? I have to say that there are no great pre-clinical models.

The whole idea of adding ipilimumab there was that CTLA-4 induces an immune response in a different way from PD-1. It's synergistic in many other tumor types, it appears to have a little bit higher response in the MSI-high colorectal cancer. Why not give it a chance and add a CTLA-4 inhibitor to a PD-1 inhibitor in the setting of a TKI and see whether we can accentuate the responses?

We did a single-center clinical trial with 39 patients. Twenty-nine patients were treated at the recommended phase 2 dose. In those 29 patients, we enriched those with nonliver mets, and seven patients with liver mets had no response. Twenty-two patients without liver metastases had an objective response rate of 36%. When we look at iRECIST, we had a 40% response rate. Of those 22 patients, six patients had no progressive disease at the 2-year mark. Some of these patients are being followed, again without any therapy, without any evidence of disease progression. These are things we haven't seen with rego-nivo before to that level, and we haven't seen with PD-1 inhibitors alone.

We know from prior data with tremelimumab monotherapy doesn't also work in colorectal cancer. We do believe there's synergy, and that's in line with the additional data that have been shown by Agenus on the botensilimab-balstilimab combination.

Schlechter: That brings us to a great next topic, which is botensilimab and balstilimab. This is the second-generation CTLA-4 inhibitor with a PD-1 inhibitor from Agenus here in Boston. You participated in that phase 1 trial, as did I. And you're now leading enrollment of the phase 2 trial.

Talk to me a little about your experience with that combination and the toxicity profile, the efficacy profile, and how it compares with regorafenib with ipilimumab and nivolumab. I'm curious to hear your thoughts.

Fakih: Botensilimab, as you mentioned, is a second-generation CTLA-4 Fc-engineered agent, and it appears to be more potent in stimulating the immune system than are first-generation CTLA-4 inhibitors. There's been no report previously on a combination of a CTLA-4, PD-1, or CTLA-4 programmed death ligand 1 (PD-L1) inhibitor with such robust responses. I think this could be potentially game-changing. The response rates reported at a couple of national meetings now are more than 30% in patients without liver metastases. The median progression-free survival of patients without liver metastases has not been reported yet on this study, with many patients without liver metastases having ongoing objective responses without progression more than a year after initiating therapy.

I think we're looking at a different class of CTLA-4 inhibitor here that may be potentially game-changing for patients with microsatellite-stable disease. It also confirms our prior findings that liver metastases are different, and we've shown that the tumor microenvironment in liver mets is completely different from that of lung mets. It has a different infiltration, B cells and T cells, and spatial distribution. That's good to know, because now you use the drugs that work in the patients where you expect those drugs to work and you focus on the liver metastases and understanding how to target those in a different way. I'm excited about that. It does appear to have colitis as an associated toxicity, but there are ways to mitigate those or to address them if you catch them early. I think a lot of that is happening now in the phase 2 trial. The phase 2 trial will be a key study to confirm a differentiation between these two drugs and support a much higher efficacy than what's available as far as standard of care.

Schlechter: For our listeners, if you are treating a patient and you don't have access either to the botensilimab-balstilimab trial or other similar investigations: If you have, say, a fit patient with no liver metastases — only peritoneal or lung disease, for example — or nodal disease that's nonresectable, do you think there's a role for off-label use of regorafenib-ipilimumab-nivolumab (RIN)? Is this something that people should be pursuing? Or is it not ready for prime time? It's always a hard question to ask, about doing things off-label. I'm curious to hear your thoughts.

Fakih: I personally think those patients should seek a trial. I think that RIN is an effective regimen, but the reality is it's still a small number of patients. It's a single-institution trial, and these patients can still have toxicities. I think we need to collect those data. We presented data at the American Association for Cancer Research meeting that looked at patients treated with peritoneal carcinomatosis who were treated RIN and rego-nivo, we didn't see a response. Granted, it was only about seven patients, but that's why it's very important to understand the pattern of metastatic disease.

I think lung mets are going to respond differently from soft-tissue metastases, which respond differently again from peritoneal carcinomatosis. We're doing our best to report the data that we have collected on almost 100 patients treated with TKI/PD-1 combinations or TKIs, PD-1 inhibitors, and CTLA-4 inhibitors. I would personally, even in my practice, enroll those patients on study.

Schlechter: What's the path forward now, do you think, with immunotherapy? Are we going to be focusing on the addition of liver-directed therapy? What other agents or novel interventions do you see over the next few years?

Fakih: For liver metastases, I think the question is going to be, is it oligometastatic disease? Is it diffuse metastatic disease? Once we confirm that the extrahepatic metastatic disease works beautifully with a certain IO combination, the next question is going to be, if you have liver limited disease, can you take care of it surgically or by ablation? If it's only two lesions and this patient is rendered NED (no evidence of disease) in the liver, could they be a candidate for IO as well? There's no question that we've seen patients with hepatectomy who have beautiful durable responses with RIN and some, frankly, with even with rego-nivo. But I don't think that if you have somebody with 10 lesions on the liver, you're suddenly going to change the biology of liver metastases. There is a lot more to understand as to what to do with those patients.

At the same time, there are other immunotherapies. You've got CAR T-cells coming online. You've got T-cell receptors coming online. I think those may also be a game changer hopefully for metastatic microsatellite-stable colorectal cancer, maybe irrespective of the location of the disease. We've got to look at all those tools.

Schlechter: That's fantastic, and thank you for that insight. Obviously, we're all optimistic about this. There's a lot to discover. I'm looking forward to reading what you have to write and listening to what you have to say about this over the coming years. That was a really great discussion.

Today, we discussed the use of immune checkpoint blockade in colorectal cancer with our guest, Dr Marwan Fakih. We talked about strategies to use this in localized rectal cancer, where it seems like there was a practice-changing result with the data on dostarlimab and the emerging data on neoadjuvant therapy in the NICHE study with ipilimumab and nivolumab. We talked about the use of pembrolizumab vs ipilimumab with nivolumab in advanced mismatch repair–deficient disease. We touched on emerging data using dual checkpoint blockade with either TKIs or novel second-generation CTLA-4 inhibitors. We're waiting for mature data on that, but we think a lot is coming and it's an exciting time.

Thank you so much for joining us. This is Dr Benjamin Schlechter for InDiscussion.


Colorectal Cancer

Microsatellite Instability in Colorectal Cancer: Overview of Its Clinical Significance and Novel Perspectives

PD-1 Blockade in Mismatch Repair-Deficient, Locally Advanced Rectal Cancer

Dostarlimab: The Miracle Drug for the Treatment of Colorectal Cancer

Neoadjuvant Immunotherapy Leads to Pathological Responses in MMR-Proficient and MMR-Deficient Early-Stage Colon Cancers

Lynch Syndrome

Optimal Management of Low Anterior Resection Syndrome: Colorectal Cancer (Update)

Pembrolizumab (Rx)


Association of Tumour Mutational Burden With Outcomes in Patients With Advanced Solid Tumours Treated With Pembrolizumab: Prospective Biomarker Analysis of the Multicohort, Open-Label, Phase 2 KEYNOTE-158 Study

Pembrolizumab Versus Chemotherapy for Microsatellite Instability-High or Mismatch Repair-Deficient Metastatic Colorectal Cancer (KEYNOTE-177): Final Analysis of a Randomised, Open-Label, Phase 3 Study

Circulating Tumor DNA Test May Help Guide GI Cancer Management

Durvalumab and Tremelimumab Combination Therapy Versus Durvalumab or Tremelimumab Monotherapy for Patients With Solid Tumors: A Systematic Review and Meta-Analysis

FOLFOXIRI/Bevacizumab Plus Nivolumab as First-Line Treatment in Metastatic Colorectal Cancer RAS/BRAF Mutated: Safety Run-In of Phase II NIVACOR Trial

Use of Total Neoadjuvant Therapy for Locally Advanced Rectal Cancer: Initial Results From the Pembrolizumab Arm of a Phase 2 Randomized Clinical Trial

Regorafenib, Ipilimumab, and Nivolumab for Patients With Microsatellite Stable Colorectal Cancer and Disease Progression With Prior Chemotherapy: A Phase 1 Nonrandomized Clinical Trial

Regorafenib Plus Nivolumab in Patients With Advanced Gastric or Colorectal Cancer: An Open-Label, Dose-Escalation, and Dose-Expansion Phase Ib Trial (REGONIVO, EPOC1603)

Regorafenib Plus Nivolumab in Patients With Mismatch Repair-Proficient/Microsatellite Stable Metastatic Colorectal Cancer: A Single-Arm, Open-Label, Multicentre Phase 2 Study

CTLA-4 Cytotoxic T-Lymphocyte Associated Protein 4 [ Homo sapiens (human) ]

iRECIST: Guidelines for Response Criteria for Use in Trials Testing Immunotherapeutics

Fc-Engineered Anti-CTLA-4 Monoclonal Antibody in Advanced Cancer

A Study of Botensilimab and Balstilimab for the Treatment of Colorectal Cancer

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.